KRISTINA at the 26th International Conference on Computational Linguistics (COLING 2016)
Oct. 1, 2016, noon
One of the KRISTINA’s scientific objectives is to research and develop technologies for generation of expressive communication. KRISTINA is going to be a virtual character able to assist elderly people and migrants with language barriers. An expressive multilingual discourse is one of the major challenges.
With this objective a new research paper has been accepted at the 26th International Conference on Computational Linguistics COLING 2016, at Osaka, Japan, from 11th to 16th December 2016. All participants were invited to present an original research on all aspects of computational linguistics, such as syntactic and semantic, dialogue and conversational agents, language generation…
This paper has been submitted by the TALN Reseach Group at the Universidad Pompeu Fabra.
Our partners Mónica Domínguez, Mireia Farrús, Leo Wanner will present their research on “An Automatic Prosody Tagger for Spontaneous Speech”.
Speech prosody is known to be central in advanced communication technologies. However, despite the advances of theoretical studies in speech prosody, so far, no large scale prosody annotated resources that would facilitate empirical research and the development of empirical computational approaches are available. This is to a large extent due to the fact that current common prosody annotation conventions offer a descriptive framework of intonation contours and phrasing based on labels that draw upon language-dependent features such as part of speech, syntactic dependencies, and punctuation, as well as on words. Prosodic units in terms of prosodic words and prosodic phrases, which are central especially in spontaneous monologues and dialogues, are not taken into account. This makes it difficult to reach a satisfactory inter-annotator agreement during the annotation of gold standard annotations and, subsequently, to create consistent large scale annotations. To address this problem, we present an annotation schema for prominence and boundary labeling of prosodic phrases based upon acoustic (instead of linguistic) parameters and a tagger for prosody annotation at the prosodic phrase level. Evaluation proves that inter-annotator agreement reaches satisfactory thresholds, from 0.60 to 0.80 Cohen’s kappa, and the prosody tagger achieves an accuracy above 0.89 for five spontaneous samples used in the evaluation of monologue and dialogue formats in English and Spanish.
See you in Osaka!